INDIAN INSTITUTE OF MANAGEMENT CALCUTTA WORKING PAPER SERIES WPS No. 569/ November 2005 A Continuous Time State Task Network model for Short Term Scheduling of Operations in Cascaded Continuous Processing Units by

In this paper we propose a novel continuous time model based on state task network for the problem of scheduling m products in n continuous processing units with simultaneous arrival of liquid inputs and multiple intermediate upliftment dates. The units can process only one product line at a time and are buffered by fixed capacity storage tanks. An optimization model for the problem tries to balance the spillage cost, upliftment failure penalty and the changeover cost. The present work develops a continuous time MINLP model for the problem based on the global event point representation scheme. The model is more general as we do not pre-fix some of the event point variables on the dates of upliftments. The MINLP model is linearized using standard linearization techniques and the resulting MILP model can output, in reasonable time, optimal schedule for 2 weeks for a 3-unit 3-product scenario as encountered by us in a refinery.